Engram, a company focused on artificial intelligence infrastructure, has secured $98 million in funding. The investment will go toward improving how AI models run — making them faster, cheaper, and less power-hungry. The round closed this week, the company confirmed, without disclosing the investors.
The efficiency challenge
Large AI models require enormous computing power. Training a single model can cost millions of dollars in electricity and hardware. Once deployed, running those models for users adds even more costs. The inefficiency has become a bottleneck for companies trying to scale AI products.
Engram’s approach targets the software layer that sits between the AI model and the hardware. By optimizing the way models are compiled, the company says it can cut inference time by a significant margin without sacrificing accuracy. That means a chatbot or image generator could respond faster on the same servers — or a company could serve more users with fewer machines.
Funding for optimization
The $98 million raise gives Engram runway to expand its engineering team and accelerate product development. The company plans to hire engineers specialized in compiler design, machine learning systems, and low-level hardware optimization.
Investors see a growing market: businesses that deploy AI want to control operational costs. As models grow larger — and as energy prices remain volatile — efficiency software becomes a competitive advantage. Engram’s funding suggests confidence that the market for model optimization tools will expand rapidly.
What comes next
Engram has not named specific benchmarks or a release date for its next product update. The company said only that the capital will allow it to “push the boundaries of what’s possible” in AI efficiency. For now, the focus is on hiring and research.
Whether Engram can deliver on its promise — and whether customers will see real savings — remains an open question. The company has the money. Now it has to build.
